CN109241951A - Porny recognition methods, identification model construction method and identification model and computer readable storage medium - Google Patents
Porny recognition methods, identification model construction method and identification model and computer readable storage medium Download PDFInfo
- Publication number
- CN109241951A CN109241951A CN201811258786.1A CN201811258786A CN109241951A CN 109241951 A CN109241951 A CN 109241951A CN 201811258786 A CN201811258786 A CN 201811258786A CN 109241951 A CN109241951 A CN 109241951A
- Authority
- CN
- China
- Prior art keywords
- human body
- pixel
- porny
- picture
- human
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/26—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Human Computer Interaction (AREA)
- Image Analysis (AREA)
Abstract
The present invention provides a kind of porny recognition methods, identification model construction method and identification model and computer readable storage mediums, receive picture to be identified；Human body pixel segmentation is carried out to picture is received, is partitioned into human body pixel region；Pornographic human bioequivalence is carried out to the human body pixel region being partitioned into, exports recognition result；On the basis of semantic segmentation network, classification described in each pixel is recovered from the abstract characteristics of convolutional layer；It uses warp lamination to up-sample receptive field for the characteristic pattern of 8 convolutional layer, the convolutional layer of the receptive field 8 is made to be restored to the identical size of input picture；Classified pixel-by-pixel on the characteristic pattern of the up-sampling, distinguishes human body pixel；The classification includes human body and non-human classification.Compared with prior art, technical solution of the present invention can reduce algorithm identification range, improve porny recognition accuracy and efficiency.
Description
Technical field
The present invention relates to image domains, in particular to a kind of porny recognition methods, identification model construction method and knowledge
Other model and computer readable storage medium.
Background technique
With the fast development of internet and multimedia technology, image and video have become main information carrier.Interconnection
The opening and interactivity of net height, have greatly facilitated cultural spreading activity, but also produce illegal erotic activity therewith.Repeatly
Prohibiting more than porn site not only polluted network environment, also seriously affect the physical and mental health of minor.Rising in recent years
Network direct broadcasting, cloud disk etc. also become the tool that criminal propagates obscene pornography, therefore the detection of Pornograph is increasingly
By the concern for the law enforcement agency and research institution for being engaged in network security technology, the identification of obscene pornographic image is to study both at home and abroad
Persons are attempting always the major issue solved in decades.
Image-recognizing method is generally divided into the image recognition side based on traditional machine learning method and based on deep learning
Method.Pornographic image identification is similar with general pattern recognition methods, pornographic image particular attribute itself is different from, such as the large area colour of skin
Exposed, pornographic movement, sensitive part are exposed etc..Identify have based on skin based on traditional machine learning method for pornographic image
The bad image recognition of color detection, the bad image recognition of view-based access control model bag of words algorithm (Bag-of-visual-words, BoVW)
Deng；The mistake that pornographic image identification based on deep learning has Yahoo to be based on 50 layers of residual error learning network (Residual Network)
Filter NSFW (Not Suitable/Safe For Work) image.Since the main body personage colour of skin is different in pornographic image, posture is more
The features such as change, pornographic image filtering are extremely difficult to 100% filtering, therefore the emphasis of always industry research.
Summary of the invention
The present invention provides a kind of porny recognition methods, identification model construction method and identification models, have and improve
The characteristics of porny recognition efficiency.
The present invention also provides a kind of computer readable storage medium, having being capable of above-mentioned any method convenient to carry out
The characteristics of.
A kind of porny recognition methods provided according to the present invention, including,
Receive picture to be identified；Human body pixel segmentation is carried out to picture is received, is partitioned into human body pixel region；To being partitioned into
Human body pixel region carry out pornographic human bioequivalence, export recognition result.
The method also includes to non-human pixel region filling 0, obtaining pure after carrying out human body pixel region segmentation
The picture of human body.
The method also includes first judging whether there is human body appearance, if it is not, then directly to received picture to be identified
Export recognition result；If it is, carrying out human body pixel segmentation again.
Human body pixel segmentation specific method include,
On the basis of semantic segmentation network, class described in each pixel is recovered from the abstract characteristics of convolutional layer
Not；It uses warp lamination to up-sample receptive field for the characteristic pattern of 8 convolutional layer, makes the convolutional layer of the receptive field 8
It is restored to the identical size of input picture；Classified pixel-by-pixel on the characteristic pattern for up-sampling and obtaining, distinguishes human body
Pixel；The classification includes human body and non-human classification.
The semantic segmentation network being based on is FCN semantic segmentation network.
A kind of porny identification model construction method provided according to the present invention, is realized based on the above method,
Including,
A large amount of pornographic human body pictures and non-pornographic human body picture are collected, and to pornographic human body classification and non-pornographic human body classification
It is labeled；Proportionally it is divided into training set, verifying collection and test set；
Construct human body pixel divide network, using human body pixel divide method extract the training set, verifying collection and
The human body pixel region of test set；
Construct the pornographic human bioequivalence network based on the human body pixel region being partitioned into, by repeatedly training, verifying and
Test finally obtains and meets the porny identification model that setting identifies accurate required precision.
A kind of porny identification model provided according to the present invention, including,
Picture receiving interface receives picture to be identified；
Human body pixel divides module, is split to human body pixel, extracts the human body pixel region of picture to be identified；
Pornographic human bioequivalence module, the human body pixel region based on picture to be identified carry out pornographic human bioequivalence and export
Recognition result；
Porny recognition result output interface exports porny recognition result.
The human body pixel segmentation module includes non-human 0 fills unit of pixel region, is filled to non-human pixel region
0, obtain the picture of pure human body.
Further include human body picture judgment module, to received picture to be identified, first judges whether there is human body appearance, if
It is no, then directly export recognition result；If it is, carrying out human body pixel segmentation again.
A kind of computer readable storage medium provided according to the present invention, be stored with load and execute convenient for processor it is above-mentioned
The computer program of the method for any one.
Compared with prior art, technical solution of the present invention can reduce algorithm identification range, and it is quasi- to improve porny identification
True rate and efficiency.
Specific embodiment
In order to make the objectives, technical solutions, and advantages of the present invention clearer, with reference to embodiments, to the present invention
It is further elaborated.It should be appreciated that described herein, specific examples are only used to explain the present invention, is not used to limit
The fixed present invention.
Any feature disclosed in this specification (including abstract) unless specifically stated can or tool equivalent by other
There are the alternative features of similar purpose to be replaced.That is, unless specifically stated, each feature is a series of equivalent or similar characteristics
In an example.
A kind of porny recognition methods provided according to the present invention, including,
Receive picture to be identified；Human body pixel segmentation is carried out to picture is received, is partitioned into human body pixel region；To being partitioned into
Human body pixel region carry out pornographic human bioequivalence, export recognition result.
The present invention proposes a kind of porny recognition methods based on human body attention mechanism, analyzes porny feature,
Human body segmentation first is carried out to the picture to be analyzed, human body image region is obtained, further human body image region is analyzed
Identification, obtain whether be porny recognition result, by it is traditional based on global image identification technical solution, be converted into specially
The image recognition in human body is infused, the conventional thought of the prior art has been broken.Technical solution of the present invention can reduce algorithm identification
Range improves porny recognition accuracy and efficiency.
The method also includes to non-human pixel region filling 0, obtaining pure after carrying out human body pixel region segmentation
The picture of human body.
As one embodiment of the present invention, after carrying out human body pixel region segmentation, to non-human pixel region background
0 processing is filled out, pure human body image is first obtained, all background pixel interference is removed, it is quasi- to further improve porny identification
True rate and efficiency.
As one embodiment of the present invention, the method also includes first judging whether to received picture to be identified
There is human body appearance, if it is not, then directly exporting recognition result；If it is, carrying out human body pixel segmentation again, further improve
Porny recognition efficiency.
Human body pixel segmentation specific method include,
On the basis of semantic segmentation network, class described in each pixel is recovered from the abstract characteristics of convolutional layer
Not；It uses warp lamination to up-sample receptive field for the characteristic pattern (feature map) of 8 convolutional layer, makes the impression
Open country is restored to the identical size of input picture for 8 convolutional layer；Divided pixel-by-pixel on the characteristic pattern for up-sampling and obtaining
Class distinguishes human body pixel；The classification includes human body and non-human classification.
As one embodiment of the present invention, human body semanteme is detected using semantic segmentation mode, and non-human example, energy
Connecting each other for human body in more people's images is enough further contemplated, the attribute of image is judged by human body relationship.
The semantic segmentation network being based on is FCN semantic segmentation network, is recovered from the abstract characteristics of convolutional layer each
Classification described in pixel further extends into the classification of pixel scale from the classification of image level；FCN can receive any ruler
Very little input picture uses warp lamination to up-sample receptive field for the characteristic pattern (feature map) of 8 convolutional layer,
The convolutional layer of the receptive field 8 is set to be restored to the identical size of input picture, so as to produce one to each pixel
A prediction, while remaining the spatial information in original input picture；It is finally enterprising in the characteristic pattern obtained in the up-sampling
Row is classified pixel-by-pixel, distinguishes human body pixel.
A kind of porny identification model construction method provided according to the present invention, is realized based on the above method,
Including,
A large amount of pornographic human body pictures and non-pornographic human body picture are collected, and to pornographic human body classification and non-pornographic human body classification
It is labeled；Proportionally it is divided into training set, verifying collection and test set；
Construct human body pixel divide network, using human body pixel divide method extract the training set, verifying collection and
The human body pixel region of test set；
Construct the pornographic human bioequivalence network based on the human body pixel region being partitioned into, by repeatedly training, verifying and
Test finally obtains and meets the porny identification model that setting identifies accurate required precision.
Pornographic image identification is absorbed in human bioequivalence by technical solution of the present invention, is reduced complex background influence, is reduced
Network training difficulty.
As embodiments of the present invention, on the basis of the data set based on acquisition, human body pixel region is extracted, then be based on
Database training training convolutional neural networks (such as ResNet50 network) obtain pornographic human body and non-pornographic human body disaggregated model.
A kind of porny identification model provided according to the present invention, including,
Picture receiving interface receives picture to be identified；
Human body pixel divides module, is split to human body pixel, extracts the human body pixel region of picture to be identified；
Pornographic human bioequivalence module, the human body pixel region based on picture to be identified carry out pornographic human bioequivalence and export
Recognition result；
Porny recognition result output interface exports porny recognition result.
The human body pixel segmentation module includes non-human 0 fills unit of pixel region, is filled to non-human pixel region
0, obtain the picture of pure human body.
Further include human body picture judgment module, to received picture to be identified, first judges whether there is human body appearance, if
It is no, then directly export recognition result；If it is, carrying out human body pixel segmentation again.
As one embodiment of the present invention, the pornographic image testing process based on human body attention mechanism is, to be measured
Image first inputs in semantic segmentation network FCN, if there is human body appearance in image, exports human region and carries on the back to inhuman body region
Scape fills out 0 processing；Human region image is sent into the convolutional neural networks ResNet50 that can recognize pornographic human body, judges that human body is
Pornographic human body or normal human.If the output of FCN network is free of human region, testing image is normal picture.
A kind of computer readable storage medium provided according to the present invention, be stored with load and execute convenient for processor it is above-mentioned
The computer program of the method for any one.
Claims (10)
1. a kind of porny recognition methods, including,
Receive picture to be identified；Human body pixel segmentation is carried out to picture is received, is partitioned into human body pixel region；To the people being partitioned into
Volumetric pixel region carries out pornographic human bioequivalence, exports recognition result.
2. porny recognition methods according to claim 1, the method also includes carrying out human body pixel region segmentation
Afterwards, to non-human pixel region filling 0, the picture of pure human body is obtained.
3. porny recognition methods according to claim 1, the method also includes, to received picture to be identified,
Human body appearance is first judged whether there is, if it is not, then directly exporting recognition result；If it is, carrying out human body pixel segmentation again.
4. according to claim 1 to porny recognition methods described in one of 3, the specific method of human body pixel segmentation includes,
On the basis of semantic segmentation network, classification described in each pixel is recovered from the abstract characteristics of convolutional layer；It adopts
Characteristic pattern with warp lamination to receptive field for 8 convolutional layer up-samples, and is restored to the convolutional layer of the receptive field 8
The identical size of input picture；Classified pixel-by-pixel on the characteristic pattern for up-sampling and obtaining, distinguishes human body pixel；Institute
Stating classification includes human body and non-human classification.
5. porny recognition methods according to claim 4, the semantic segmentation network being based on is FCN semantic segmentation net
Network.
6. a kind of porny identification model construction method, method described in one of claims 1 to 5 on the basis of, is realized,
Including,
A large amount of pornographic human body pictures and non-pornographic human body picture are collected, and pornographic human body classification and non-pornographic human body classification are carried out
Mark；Proportionally it is divided into training set, verifying collection and test set；
It constructs human body pixel and divides network, the training set is extracted using the method that human body pixel is divided, verifying collects and test
The human body pixel region of collection；
The pornographic human bioequivalence network based on the human body pixel region being partitioned into is constructed, by training, verifying and test repeatedly,
It finally obtains and meets the porny identification model that setting identifies accurate required precision.
7. a kind of porny identification model, which is characterized in that including,
Picture receiving interface receives picture to be identified；
Human body pixel divides module, is split to human body pixel, extracts the human body pixel region of picture to be identified；
Pornographic human bioequivalence module, the human body pixel region based on picture to be identified carry out pornographic human bioequivalence and export identification
As a result；
Porny recognition result output interface exports porny recognition result.
8. porny identification model according to claim 7, which is characterized in that the human body pixel divides module and includes
Non-human 0 fills unit of pixel region obtains the picture of pure human body to non-human pixel region filling 0.
9. porny identification model according to claim 7 or 8, which is characterized in that further include that human body picture judges mould
Block first judges whether there is human body appearance to received picture to be identified, if it is not, then directly exporting recognition result；If so,
Human body pixel segmentation is then carried out again.
10. a kind of computer readable storage medium, which is characterized in that be stored with convenient for processor load and perform claim requires 1
To the computer program of the method for any one in 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811258786.1A CN109241951A (en) | 2018-10-26 | 2018-10-26 | Porny recognition methods, identification model construction method and identification model and computer readable storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811258786.1A CN109241951A (en) | 2018-10-26 | 2018-10-26 | Porny recognition methods, identification model construction method and identification model and computer readable storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109241951A true CN109241951A (en) | 2019-01-18 |
Family
ID=65082452
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811258786.1A Pending CN109241951A (en) | 2018-10-26 | 2018-10-26 | Porny recognition methods, identification model construction method and identification model and computer readable storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109241951A (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110414519A (en) * | 2019-06-27 | 2019-11-05 | 众安信息技术服务有限公司 | A kind of recognition methods of picture character and its identification device |
| CN110555482A (en) * | 2019-09-06 | 2019-12-10 | 腾讯科技（深圳）有限公司 | Vulgar picture identification method and device based on artificial intelligence and electronic equipment |
| CN110807139A (en) * | 2019-10-23 | 2020-02-18 | 腾讯科技（深圳）有限公司 | Picture identification method and device, computer readable storage medium and computer equipment |
| CN111178343A (en) * | 2020-04-13 | 2020-05-19 | 腾讯科技（深圳）有限公司 | Multimedia resource detection method, device, equipment and medium based on artificial intelligence |
| CN111199233A (en) * | 2019-12-30 | 2020-05-26 | 四川大学 | Improved deep learning pornographic image identification method |
| CN112016345A (en) * | 2019-05-28 | 2020-12-01 | 腾讯科技（深圳）有限公司 | Image identification method and device |
| CN115546824A (en) * | 2022-04-18 | 2022-12-30 | 荣耀终端有限公司 | Taboo picture identification method, equipment and storage medium |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104813339A (en) * | 2012-09-12 | 2015-07-29 | 威智伦富智堡公司 | Methods, devices and systems for detecting objects in a video |
| CN106570869A (en) * | 2016-10-17 | 2017-04-19 | 安徽慧视金瞳科技有限公司 | Poisson equation and gradient-based human body image segmentation method |
| WO2017101036A1 (en) * | 2015-12-16 | 2017-06-22 | Intel Corporation | Fully convolutional pyramid networks for pedestrian detection |
| CN106951848A (en) * | 2017-03-13 | 2017-07-14 | 平安科技（深圳）有限公司 | The method and system of picture recognition |
| CN107424159A (en) * | 2017-07-28 | 2017-12-01 | 西安电子科技大学 | Image, semantic dividing method based on super-pixel edge and full convolutional network |
| CN107516316A (en) * | 2017-07-19 | 2017-12-26 | 郑州禅图智能科技有限公司 | It is a kind of that the method that focus mechanism is split to static human image is introduced in FCN |
| CN107665333A (en) * | 2017-08-28 | 2018-02-06 | 平安科技（深圳）有限公司 | A kind of indecency image identification method, terminal, equipment and computer-readable recording medium based on convolutional neural networks |
| CN108198192A (en) * | 2018-01-15 | 2018-06-22 | 任俊芬 | A kind of quick human body segmentation's method of high-precision based on deep learning |
| CN108229479A (en) * | 2017-08-01 | 2018-06-29 | 北京市商汤科技开发有限公司 | The training method and device of semantic segmentation model, electronic equipment, storage medium |
-
2018
- 2018-10-26 CN CN201811258786.1A patent/CN109241951A/en active Pending
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104813339A (en) * | 2012-09-12 | 2015-07-29 | 威智伦富智堡公司 | Methods, devices and systems for detecting objects in a video |
| WO2017101036A1 (en) * | 2015-12-16 | 2017-06-22 | Intel Corporation | Fully convolutional pyramid networks for pedestrian detection |
| CN106570869A (en) * | 2016-10-17 | 2017-04-19 | 安徽慧视金瞳科技有限公司 | Poisson equation and gradient-based human body image segmentation method |
| CN106951848A (en) * | 2017-03-13 | 2017-07-14 | 平安科技（深圳）有限公司 | The method and system of picture recognition |
| CN107516316A (en) * | 2017-07-19 | 2017-12-26 | 郑州禅图智能科技有限公司 | It is a kind of that the method that focus mechanism is split to static human image is introduced in FCN |
| CN107424159A (en) * | 2017-07-28 | 2017-12-01 | 西安电子科技大学 | Image, semantic dividing method based on super-pixel edge and full convolutional network |
| CN108229479A (en) * | 2017-08-01 | 2018-06-29 | 北京市商汤科技开发有限公司 | The training method and device of semantic segmentation model, electronic equipment, storage medium |
| CN107665333A (en) * | 2017-08-28 | 2018-02-06 | 平安科技（深圳）有限公司 | A kind of indecency image identification method, terminal, equipment and computer-readable recording medium based on convolutional neural networks |
| CN108198192A (en) * | 2018-01-15 | 2018-06-22 | 任俊芬 | A kind of quick human body segmentation's method of high-precision based on deep learning |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112016345A (en) * | 2019-05-28 | 2020-12-01 | 腾讯科技（深圳）有限公司 | Image identification method and device |
| CN110414519A (en) * | 2019-06-27 | 2019-11-05 | 众安信息技术服务有限公司 | A kind of recognition methods of picture character and its identification device |
| CN110414519B (en) * | 2019-06-27 | 2023-11-14 | 众安信息技术服务有限公司 | Picture character recognition method and device and storage medium |
| CN110555482A (en) * | 2019-09-06 | 2019-12-10 | 腾讯科技（深圳）有限公司 | Vulgar picture identification method and device based on artificial intelligence and electronic equipment |
| CN110807139A (en) * | 2019-10-23 | 2020-02-18 | 腾讯科技（深圳）有限公司 | Picture identification method and device, computer readable storage medium and computer equipment |
| CN110807139B (en) * | 2019-10-23 | 2023-09-01 | 腾讯科技（深圳）有限公司 | Picture identification method, device, computer readable storage medium and computer equipment |
| CN111199233A (en) * | 2019-12-30 | 2020-05-26 | 四川大学 | Improved deep learning pornographic image identification method |
| CN111199233B (en) * | 2019-12-30 | 2020-11-20 | 四川大学 | Improved deep learning pornographic image identification method |
| CN111178343A (en) * | 2020-04-13 | 2020-05-19 | 腾讯科技（深圳）有限公司 | Multimedia resource detection method, device, equipment and medium based on artificial intelligence |
| CN115546824A (en) * | 2022-04-18 | 2022-12-30 | 荣耀终端有限公司 | Taboo picture identification method, equipment and storage medium |
| CN115546824B (en) * | 2022-04-18 | 2023-11-28 | 荣耀终端有限公司 | Taboo picture identification method, apparatus and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109241951A (en) | Porny recognition methods, identification model construction method and identification model and computer readable storage medium | |
| Tian et al. | A dual neural network for object detection in UAV images | |
| CN105488517B (en) | A kind of vehicle brand type identifier method based on deep learning | |
| CN115063573B (en) | Multi-scale target detection method based on attention mechanism | |
| CN110349136A (en) | A kind of tampered image detection method based on deep learning | |
| CN102867188B (en) | Method for detecting seat state in meeting place based on cascade structure | |
| CN115082855B (en) | Pedestrian shielding detection method based on improved YOLOX algorithm | |
| CN107085696A (en) | A kind of vehicle location and type identifier method based on bayonet socket image | |
| CN107122375A (en) | The recognition methods of image subject based on characteristics of image | |
| CN106650806A (en) | Cooperative type deep network model method for pedestrian detection | |
| CN109657551B (en) | Face detection method based on context information enhancement | |
| CN103366160A (en) | Objectionable image distinguishing method integrating skin color, face and sensitive position detection | |
| CN109948616A (en) | Image detecting method, device, electronic equipment and computer readable storage medium | |
| CN109145745A (en) | A kind of face identification method under circumstance of occlusion | |
| CN104134059A (en) | Color-information-maintaining objectionable image detection method under deformation sensitive organ models | |
| CN108734138A (en) | A kind of melanoma skin disease image classification method based on integrated study | |
| CN105117708A (en) | Facial expression recognition method and apparatus | |
| CN106203296B (en) | The video actions recognition methods of one attribute auxiliary | |
| CN103440494A (en) | Horrible image identification method and system based on visual significance analyses | |
| CN109255319A (en) | For the recognition of face payment information method for anti-counterfeit of still photo | |
| CN108257151A (en) | PCANet image change detection methods based on significance analysis | |
| CN110059541A (en) | A kind of mobile phone usage behavior detection method and device in driving | |
| CN112949469A (en) | Image recognition method, system and equipment for face tampered image characteristic distribution | |
| Mack et al. | Modeling categorization of scenes containing consistent versus inconsistent objects | |
| CN114677707A (en) | Human body posture estimation method based on multi-resolution feature fusion network |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| EE01 | Entry into force of recordation of patent licensing contract | ||
| EE01 | Entry into force of recordation of patent licensing contract |
Application publication date: 20190118 Assignee: Apple R&D (Beijing) Co., Ltd. Assignor: BEIJING MOSHANGHUA TECHNOLOGY CO., LTD. Contract record no.: 2019990000054 Denomination of invention: Porny recognition methods, identification model construction method and identification model and computer readable storage medium License type: Exclusive License Record date: 20190211 |